AIAA 98-3783

CHARGES IN PYROTECHNIC
MECHANISMS

Laurence J. Bement
NASA Langley Research Center
Hampton, VA

34th AIAA/ASME/SAE/ASEE
Joint Propulsion Conference & Exhibit
July 13-15, 1998 / Cleveland, OH
For permission to copy or republish, contact the American Institute of Aeronautics and Astronautics
1801 Alexander Bell Drive, Suite 500, Reston, VA 20191

A METHOD FOR SIZING BOOSTER CHARGES IN PYROTECHNIC MECHANISMS
Laurence J. Bement *
NASA Langley Research Center
Hampton, VA
ABSTRACT
Since no generally accepted guidelines exist on sizing
booster charges to assure functional margins in
pyrotechnically actuated devices, a study was
conducted to provide an approach to meet this need.
An existing pyrovalve was modified from a single
cartridge input to a dual-cartridge input with a booster
charge. The objectives of this effort were to
demonstrate an energy-based functional margin
approach for sizing booster charges, and to determine
booster charge energy delivery characteristics in this
valve. Functional margin was demonstrated by
determining the energy required to actuate the valve
through weight drop tests for comparison to the energy
delivered by the cartridge and booster charge in firings
in the modified valve. The results of this study
indicated that this energy-based approach fully met the
study objectives, showing its usefulness for this and
possibly other pyrotechnic devices.
INTRODUCTION
In designing pyrotechnic devices, the pyrotechnic
charge must be sized to deliver more energy than is
required to assure that a functional margin (reference 1)
is achieved. In order to determine functional margin
and to accommodate known variations, described in
reference 2, in the performance of gas-generating
propellant charges, energy delivery must be measured
in the device. Two approaches are generally used for
providing energy-producing, pyrotechnic charges:
(1) one or two cartridges with the full operating charge
contained within the cartridge, and (2) one or two
initiators to ignite a separate booster charge.
Unfortunately, no generally accepted design guidelines
_____________________
*Senior Pyrotechnic Engineer, AIAA Associate Fellow
Copyright © 1998 by the American Institute of Aeronautics and
Astronautics, Inc. No copyright is asserted in the United States under
Title 17, U.S. Code. The U.S. Government has a royalty-free license
to exercise all rights under the copyright claimed herein for
Governmental Purposes. All other rights are reserved by the
copyright owner.
1
American Institute of Aeronautics and Astronautics
AIAA 98-3783
exist on how to size the propellant charge in any
configuration while assuring an acceptable functional
margin. To address this issue, a research effort was
conducted, using an Apollo pyrovalve, the same Apollo
pyrovalve with a modification, and a steel test valve.
The objectives of this study were to determine the
size of the booster charge needed to assure a functional
margin for the operation of the Apollo pyrovalve, and
to determine the energy delivery characteristics of the
booster charge.
The proven functional margin test approach in
reference 1 of comparing the energy required to
function pyrotechnic mechanisms to the energy
delivered by the pyrotechnic energy source was used in
this investigation. Weight drop tests were used to
measure the energy required to function the pyrovalve.
A steel test valve was used to determine the effects of
free volume within the valve and booster charge size,
ignition, combustion, and energy delivery. The energy
delivered by several select booster charge quantities
was measured in the modified Apollo pyrovalve.
TEST HARDWARE
This section describes the valves, the cartridges and the
booster charge material used.
Apollo pyrovalve
The “normally closed” aluminum Apollo pyrovalve is
shown in figure 1. The fluid flow through the valve is
blocked by the blind fittings. As the hot-gas output of
the cartridge forces the piston downward, the piston
blade shears off the nipples on the blind fittings (0.020-
inch wall thickness). A continuation of the stroke
aligns the port in the piston blade with the now-open
fittings to allow fluid flow. Approximately 0.4 inch of
stroke is required to fully open the valve. During this
stroke, the tapered portion of the piston body engages a
matching taper in the bore at approximately 0.2 inch.
This engagement slows and stops the piston over the
next 0.2 inch. The kinetic energy of the piston is
dissipated through friction and metal deformation of the
cylinder wall. Since the cartridge used in this
investigation was much smaller than the unit originally
used on the mission, a port adapter and plug were
employed to accommodate the large thread diameter
and to eliminate the free volume introduced between
the cartridge and piston.
Modified Apollo pyrovalve
The modified Apollo pyrovalve is shown in figure 2. A
two-cartridge steel “Y” adapter was installed in the
existing threads in the pyrovalve. Firing either or both
of the cartridges would ignite the booster charge placed
in the free volume. One of the cartridges was replaced
with a piezoelectric pressure transducer, as shown, to
monitor the working pressure within the valve as it
operated. A portion of the internal free volume was
reduced by installing the two cylinders in the cartridge
bores and by extending the plug on the piston face. A
free volume of 0.6 cubic centimeters was left to receive
the varying amounts of propellant booster charge. Each
100 mg of the propellant occupied 0.10 cc, reducing the
free volume as propellant was added. The bottoms of
the cylinders in the cartridge ports were taped over to
retain the propellant within the lower portion of the free
volume during assembly. It should be noted that the
use of the tape was a matter of convenience for this
study. Steel foils, welded to the housing would be used
in flight hardware to encapsulate and prevent
contamination of the booster charge.
Steel test valve
The steel test valve, shown in figure 3, was used to
better understand the ignition, combustion and energy
delivery of the booster charge. It was designed to
accommodate the “Y” adapter and the blind fittings to
duplicate the Apollo valve interfaces. A heavy-duty
steel housing was fabricated with no taper on the piston
or the housing bore to allow for multiple reuses. The
piston stroked approximately 0.7 inch before its bottom
shoulder engaged the blind fitting. The initial
conditions, prior to piston motion, that affected booster
charge ignition and combustion in the “Y” adapter were
the same between the two configurations; the housing
materials and volumes were the same, as well as the
initial resistance to stroke in shearing the nipples. An
electrically grounded probe was installed on the piston
blade to measure the velocity of the piston. As the
probe contacted each of the electrically charged foil
switches, an electrical impulse was recorded. The
2
American Institute of Aeronautics and Astronautics
0.140-inch stroke distance between adjacent switches,
divided by the time interval required to traverse that
distance, provided the average velocity for that specific
portion of the stroke. This velocity was used to
calculate the kinetic energy of the moving mass, which
was the piston, plug and probe. The velocity of the
mass was measured at a piston stroke displacement of
0.4 inch, which is the stroke required to open the
Apollo pyrovalve. The kinetic energy at that point was
used for performance evaluations.
Cartridge
The cartridge selected for this study was the NASA
Standard Initiator (NSI)-derived Gas Generating
Cartridge (NGGC), reference 3. This cartridge has the
same body and electrical interface as does the NSI, but,
depending on the application, has at least twice the
output. The weight of the internal charge is
130 milligrams.
Booster charge
The booster charge selected for this study was an on-
hand gun propellant, reference 4, composed of an
80/10/10 by weight mixture of nitrocellulose,
nitroglycerine and other ingredients, respectively. It
quickly generates a large volume of gases, consisting
primarily of carbon monoxide, carbon dioxide, and
water. The 0.04-inch diameter, 0.02-inch thick flakes
were poured into the free volume of the valves. This
material is not stable under thermal/vacuum conditions
and is not recommended for application to aerospace
missions.
EXPERIMENTAL AND ANALYTICAL
PROCEDURES
This study was conducted in four steps: (1) tests to
determine the energy required to function the Apollo
pyrovalve; (2) tests to evaluate the performance of
pyrotechnic charges in the steel valve; (3) tests to
evaluate the performance of pyrotechnic charges in the
Apollo pyrovalve; and (4) analysis to determine the
functional margin achieved in the modified Apollo
pyrovalve.
Tests to determine the energy required to function the
Apollo pyrovalve
Dropping a weight (reference 1) onto the valve's piston
(through an adapter pin that extends above the housing)
provided a controlled input to determine the energy
required to function the valve. The impacting weight
simulates the dynamic output of the cartridge and
booster charges. A load cell was positioned under the
pyrovalve to measure the forces required to shear the
nipples and complete the stroke to open the valve. The
weight drop height was adjusted to determine the
minimum energy required to accomplish the function.
Tests were conducted (reference 5) at incremental input
energies of 100 to 790 inch-pounds to understand the
functional mechanisms of the valve. The data collected
were the amount of stroke of the piston into the valve
body taper engagement, versus energy input. This data
was used to determine the energy induced by the output
of the cartridge and cartridge/booster combination, as
described in the subsequent section, “Tests to evaluate
the performance of pyrotechnic charges in the Apollo
pyrovalve.”
Tests to evaluate the performance of pyrotechnic
charges in the steel pyrovalve
A total of 14 functional tests, measuring velocity of the
piston, were conducted in the steel pyrovalve. Four
single-cartridge tests were conducted without the “Y”
adapter (no internal free volume) to obtain a baseline.
Ten were conducted with the “Y” adapter to better
understand the ignition and combustion characteristics
of the booster. These ten tests were conducted with the
following booster charges:
Booster charge weight Free volume
No. of tests mg cc
1 50 0.55
3 100 0.50
3 200 0.40
1 300 0.30
1 400 0.20
1 500 0.10
The kinetic energy achieved at 0.4 inch of stroke of
the piston for each of these booster charges was used as
the basis for performance comparison.
Tests to evaluate the performance of pyrotechnic
charges in the Apollo pyrovalve
One single-cartridge test was conducted in the Apollo
pyrovalve without the “Y” adapter to obtain a
performance baseline. Due to a limited number of
3
American Institute of Aeronautics and Astronautics
valve bodies, only 3 tests were conducted in the
modified Apollo pyrovalve with the “Y” adapter:
Booster charge weight Free volume
No. of tests mg cc
1 200 0.40
1 400 0.20
1 500 0.10
The mechanical energy delivered by the cartridge
and booster charge combinations was determined by
measuring the piston stroke in the taper engagement,
and using the plot of “stroke versus input energy”
derived from the weight drop tests, selecting the energy
for a given measured stroke.
Analysis of functional margin
The functional margin, or excess capability to operate,
of the Apollo pyrovalve was determined by comparing
the mechanical energy delivered by the pyrotechnic
charges to the mechanical energy required to fully open
the valve. Functional Margin (reference 1) is defined
as: Energy Delivered - Energy Required
Energy Required
Where:
Energy Delivered = the energy delivered by
pyrotechnic charges in firings in the Apollo
pyrovalve.
Energy Required = the energy required to fully
open the Apollo pyrovalve, as determined by
weight drop tests.
In order to have a functional margin, the Energy
Delivered must be greater than the Energy Required.
RESULTS
The results of this research are presented in the same
sequence as the procedures section.
Energy Required to function Apollo pyrovalve
Typical force versus time plots for two different input
energy levels from the weight drop tests (reference 5)
are shown in figure 4. The first spike is the resistance
to shearing the nipples in the blind fittings. An estimate
of the internal pyrotechnic-generated pressure needed to
shear these nipples was obtained by dividing the highest
resistive force observed, approximately 1,700 pounds
by the area of the 0.52-inch diameter piston to yield
8,000 psi. At least this much pressure is needed to
shear the nipples. The oscillatory force indicates the
engagement of the piston/cylinder tapers and stroking
of the piston until it stops. The results of the weight
drop tests (reference 5), shown in figure 5, indicate that
the stroke in the taper engagement, versus energy input,
was approximately linear. The energy required to fully
open the valve at a 0.200-inch stroke in the tapered
interface, based on a linear extrapolation, is
950 inch-pounds. Referring to the actuating forces in
figure 4, virtually all of the energy consumed in
stopping the piston occurs in the taper engagement.
The energy consumed in shearing the nipples was only
about 34 inch-pounds (0.02 inch, the wall thickness of
the nipples, multiplied by the 1,700-pound peak force).
Performance of pyrotechnic charges in the steel
pyrovalve
The kinetic energies (calculated from the measurements
of the velocity the piston achieved at a stroke of 0.4
inch), versus booster charge weight and initial free
volume are listed in table I and plotted in figure 6. In
figure 6, the 5 data points collected in firing single
cartridges, with zero free volume and without the “Y”
adapter, are shown for convenience on the ordinate.
The data with the “Y” adapter (lower curve) indicates a
linear increase in energy delivered with an increase in
booster charge and the corresponding decrease in free
volume. This kinetic energy data in the steel pyrovalve
should not be compared to the energy required to
function the Apollo pyrovalve, a different mechanical
operation. Figure 7 shows typical pressure versus time
histories produced by the different booster charges.
The 50-mg booster charge produced a pressure of
10,000 psi, which was unable to shear the nipples on
the blind fittings. Thus, a better estimation of a
threshold pressure to fail the nipples is greater than
10,000 psi, rather than the 8,000 psi estimate from the
weight drop tests. Very rapid pressure rises occurred
with larger charges. Also, the larger charges produced
wider high-pressure pulses in transferring more energy
to the piston. The times to reach the 0.7-inch limit of
piston stroke for each charge are also shown in figure 7
to assist in describing functional performance.
Performance of pyrotechnic charges in Apollo
pyrovalve - The mechanical-stroke energy data
collected in these tests are listed in table I and shown in
figure 6. The energy required to fully open the valve
4
American Institute of Aeronautics and Astronautics
(950 inch-pounds, obtained from the linearly
extrapolated weight drop test data, figure 5) is shown as
the horizontal line. The firing of the single cartridge in
the Apollo pyrovalve without the “Y” adapter is plotted
on the ordinate, again for convenience. The three
firings of the selected booster charges in the modified
Apollo pyrovalve are also shown, resulting in a non-
linear increase in energy with increase in booster charge
and a corresponding decrease in free volume. The 500
mg booster charge delivered more energy (estimated at
1,050 inch-pounds, based on a linear extrapolation of
the data in figure 5) than was required to fully open the
valve. The pressure traces recorded in these three
firings are shown in figure 8. The rapid pressure rises
and peak pressures are comparable to those measured in
the steel valve. The time for the piston to reach the limit
of its stroke is no greater than 0.5 ms., based on the
flattening of the pressure traces and on the data
collected in the steel pyrovalve. The pressure rises after
about 0.25 ms with the 400 and 500-mg booster charges
indicate that some amount of the booster charge is still
burning after the piston has stopped.
Analysis of functional margin
Of all the firings in the modified Apollo pyrovalve, the
500-mg booster charge was the only one that was able
to fully open the valve. Based on an energy delivered
of 1050 inch-pounds (extrapolation in figure 5), and an
energy required of 950 inch-pounds, the functional
margin for this firing was calculated to be 0.105.
CONCLUDING REMARKS
A research program was conducted on a modified
Apollo pyrovalve to demonstrate a method for sizing
booster charges to assure functional margins in
pyrotechnic devices. A second objective was to
determine booster charge energy delivery
characteristics in this valve. The valve was modified
from a single cartridge input to a dual-cartridge input
with a booster charge. Test and analytical methods
previously proven in evaluating other devices were
adapted to this effort. Determining the energy required
to function this pyrovalve and the energy delivered by
the booster charge allows a calculation of functional
margin. The “energy required” to function the valve
was determined through weight drop tests on the
actuating mechanism, while measuring the resistive
forces during actuation. These measurements of
resistive forces assisted in understanding the mechanics
of operation in accomplishing the function. Repetitive
test firings in a steel mockup of the pyrovalve, while
measuring the pressure within the working volume and
the kinetic energy delivered into the actuating piston,
offered an improved understanding of the ignition and
combustion of the booster charges. These tests not only
gave direction in selecting the charges to be
demonstrated in the final configuration valve, but also
allowed a reduction in the number of units tested.
A number of performance effects were observed in
this study. A rapid pressure rise and high pressures
produce the highest efficiencies in transferring the
energy from the burning booster charge to the actuating
piston. The valve functions in less than
0.5 millisecond. Mockups of pyrotechnic devices are
useful, but only tests in the final configuration hardware
should be used to define specific performance. The
cartridge used in this study (130-mg charge) had
insufficient output to function the valve. In the
modified valve, a booster charge of 200 mg was needed
to match the energy delivered in the single-cartridge
configuration, and a 500-mg charge was needed to fully
open the valve. The booster charges overcame the
adverse effect caused by the increase in the size of the
working volume pressurized by the pyrotechnic
charges. The 500-mg charge produced a functional
margin of 0.105. This pyrovalve could be redesigned to
be more energy efficient by stopping the actuating
piston after the valve fully opened, rather than during
the opening phase.
This work successfully demonstrates a method for
sizing the booster charge necessary to generate a
functional margin in a modified Apollo pyrovalve. The
5
American Institute of Aeronautics and Astronautics
method should apply to, and be a useful tool for, design
of other mechanisms.
REFERENCES
1. Bement, Laurence J. and Schimmel, Morry L.:
Determination of Pyrotechnic Functional Margin.
Presented at the 21st Annual SAFE Symposium,
November 11 to 13, 1991, Las Vegas, NV.
2. Bement, Laurence J. and Schimmel, Morry L.: A
Manual for Pyrotechnic Design, Development and
Qualification. NASA TM 110172, June 1995.
3. Bement, Laurence J.; Schimmel, Morry L. and Karp,
Harold; and Magenot, Michael C.: Development
and Demonstration of an NSI-derived Gas
Generating Cartridge (NGGC). NASA Pyrotechnic
Systems Workshop, Albuquerque, NM, February
1994.
4. Bement, Laurence J.; Holmes, Clair; McGrory,
James; and Schimmel, Morry L,: An Investigation
to Improve Quality Evaluations of Primers and
Propellant for 20mm Munitions. Presented at the
Munitions Technology Symposium IV and
Statistical Process Control Conference, Reno, NV,
February 10 - 12, 1997.
5. Bement, Laurence J.: Functional Performance of
Pyrovalves. Journal of Spacecraft and Rockets,
Vol. 34, No. 3, May-June 1997.
TABLE I. SUMMARY OF ENERGY MEASUREMENTS IN PYROVALVES

Initial Free Volume Performance

Energy Source cubic centimeters inch-pounds
Steel Apollo
Single cartridge zero 276 690
(no “Y” adapter) 185
230 - Avg. = 218
195 Std. Dev. = 49
205
Single cartridge
(with “Y” adapter
and booster, mg)
50 0.55 0
100 0.50 155
100 0.50 221 - Avg. = 172
100 0.50 140
200 0.40 320 735
200 0.40 293 - Avg. = 295
200 0.40 275
300 0.30 492
400 0.20 588 875
500 0.10 711 1,050

6
American Institute of Aeronautics and Astronautics
 Cartridge

Port Adapter
Plug (attached to piston
0.3" diameter)
0.55" dia
Piston
3.50" 0.20" stroke to taper
0.52" dia Piston Blade
0.30" dia Blind Fitting (2)

0.4" stroke to open

Nipple
Figure 1. Cross sectional view of normally closed Apollo pyrovalve. The housing and fittings are aluminum.

Cartridge
(NGGC) Pressure Transducer

Steel "Y" Adapter

Cylinder (2) (0.1" dia port)
Free Volume (Taped over bottom)
(0.6cc empty)
Plug (attached to piston,
0.3" diameter)
Apollo pyrovalve housing

Piston

Piston Blade
Blind Fitting (2)

Nipple

Figure 2. Cross sectional view of Apollo pyrovalve modified to accept two inititators. The pressure transducer was
installed in the second port to measure activation pressure.
7
American Institute of Aeronautics and Astronautics
 Presssure Transducer
Cartridge
(NGGC)

Steel "Y" Adapter Steel Test Body

Piston
Straight Bore

0.7" to fitting

Blind fitting (2)

Foil contact switches (5) Electrically grounded probe

.140" Spacing

Figure 3. Cross sectional view of steel test pyrovalve used to measure the velocity of the piston
in opening the valve.

Inch-pounds input
155 (71 in. × 2.19lb)
12 000 795 (106 in. × 7.51lb)

10 000

8 000
Force,
pounds 6 000
4 000

2 000

0 1 2 3
Time, milliseconds
Figure 4. Force versus time traces for weight drop tests on Apollo pyrovalve.

8
American Institute of Aeronautics and Astronautics
Stroke into taper engagement, inch
Stroke needed to fully open
.200
(950 inch-pounds)

.150

.100

.050

0 100 200 300 400 500 600 700 800 900 1000 1100
Energy Input, inch-pounds

Figure 5. Weight drop test stroke into the taper engagement versus energy input in the Apollo pyrovalve.

1200
Energy delivered, inch-pounds

1000 Energy required to function Apollo pyrovalve

800

600
Single cartridge steel valve
400 (0 free volume)
Steel valve ("Y" adapter)
Single cartridge Apollo pyrovalve
200 (0 free volume)
Modified Apollo pyrovalve ("Y" adapter)
0
0 100 200 300 400 500
Booster charge weight, mg
.6 .5 .4 .3 .2 .1
Free volume, cc
Figure 6. The energy delivered versus booster charge weight and free volume of aluminum and steel pyrovalves.

9
American Institute of Aeronautics and Astronautics
 50mg booster charge (0 inch-pounds, didn't shear nipples)
100mg booster charge (150 inch-pounds, 0.54 ms to stroke)
200mg booster charge (273 inch-pounds, 0.46 ms to stroke)
Operating pressure, psi × 1000
70 300mg booster charge (492 inch-pounds, 0.36 ms to stroke)
400mg booster charge (588 inch-pounds, 0.32 ms to stroke)
60 500mg booster charge (711 inch-pounds, 0.33 ms to stroke)
50

40
30

20
10

0
0 .25 .50 .75 1.00 1.25 1.50 1.75
Time, milliseconds

Figure 7. Typical presure versus time histories of the booster charges tested in the steel pyrovalve. The times listed
indicate when the piston reached 0.7 inch in its stroke.

200mg charge (735 inch-pounds)

Operating pressure, psi × 1000

70
400mg charge (875 inch-pounds)
60 500mg charge (1050 inch-pounds)

50

40
30
20
10

0
0 .25 .50 .75 1.00 1.25 1.50 1.75
Time, milliseconds

Figure 8. Pressure versus time histories of the booster charges tested in the Apollo pyrovalve.

10
American Institute of Aeronautics and Astronautics